Laptop hangs during Fedora 20 boot often, but works flawlessly once booted. Strange?

asked 2014-03-07 17:50:04 -0500

bfrink gravatar image

updated 2014-03-07 22:57:44 -0500

mether gravatar image

Interestingly, I have an old Gateway ML3109 that will boot Fedora (3.13.5-200.FC20.i686) 20 (Heisenbug) swiftly and completely only about 20% of boot attempts. Most often I have to poweroff and try again and again to get a clen boot. However, once booted everything works practically flawlessly. I run Office, Inkscape, Gimp, and Octave with no problems. The wireless works great and even video streams well. Are there any configuration settings that I can change that might skip over the boot problem?

Admittingly, my Gateway laptop is junk by today's standards, but half the fun is getting something worthless to do something worthwhile. Here are the specs.

Memory: 1.4 GiB Processor: Intel Celeron M CPU 520@1.60GHz OS Type 32 bit Graphics: Gallium 0.4 on ATI RC410 GNOME: 3.10.2 Disk: 114.7 GB Hardware name: Gateway ML3109 /, BIOS 83.09 05/25/07

On most failed boots messaging stops at: [0.683044] perf samples too long (29921 > 19230), lowering Kernel.perf_event_max_sample_rate to 7000

Sometimes the boot proceeds to "ok reached targeted system" before complaining about perf samples.

Booting is relatively fast when a succesful boot occurs, and then everything works great.

Once booted, where should I look to get hints about what the true problem is? Does anyone have any suggestions on config file settings? Any advice other than throw the junk computer away will be appreciated. I may consider electronic recycling, but just not yet. I'm using the laptop in question to ask these questions!

Thank you for your time,


edit retag flag offensive close merge delete


I have recently seen similar problems with Dell E6540 ( ) and Ubuntu 13.10. The laptop had apparently broken hardware changed: memory, motherboard, but some issues persist, just less often, and it won't start X after reboot. In the Linux world older hardware was always better (or the situation has been like that for many years). Unfortunately now Linux support for old hardware, which was once supported, worsens. Perform hardware diagnostics, google, experiment with different kernels or distributions (you may get more support from Ubuntu)

marcindulak gravatar imagemarcindulak ( 2014-03-08 10:20:47 -0500 )edit

So, the perf_event_max_sample_rate thing is just a warning. There's a better explanation in or the docs it cites, but essentially your CPU is getting overworked and the kernel is slowing things down so it doesn't get worse. This warning isn't the cause of your problem, it is an effect of the problem. Besides the obvious age of the system, I would look at disk integrity as a possible issue - ie smartctl -a /dev/sda - the rest of the journal might suggest something helpful too.

randomuser gravatar imagerandomuser ( 2014-03-08 10:46:02 -0500 )edit

Thanks to all commenting. Great suggestions and I am following all of them. I will let you know if I have any breakthroughs. I am using the laptop today with no difficulty other than the frequently hanging boot process.

bfrink gravatar imagebfrink ( 2014-03-08 16:05:20 -0500 )edit